First starting with HR data which consists of employees’ personal information and employment information.
library(ggplot2)
## Warning: package 'ggplot2' was built under R version 3.5.2
library(plotly)
##
## Attaching package: 'plotly'
## The following object is masked from 'package:ggplot2':
##
## last_plot
## The following object is masked from 'package:stats':
##
## filter
## The following object is masked from 'package:graphics':
##
## layout
library(highcharter)
## Warning: package 'highcharter' was built under R version 3.5.2
## Highcharts (www.highcharts.com) is a Highsoft software product which is
## not free for commercial and Governmental use
library(dplyr)
## Warning: package 'dplyr' was built under R version 3.5.2
##
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
##
## filter, lag
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
#importing files I will work with
hr<-read.csv("~/Desktop/CSDA-1050F18S1/eugenepark/CSDA1050HR.csv")
#Examine the HR dataset
#S.EMP. - A unique employee number assigned to each employee
#C.LEVEL - Employee's job level ranging from C1 being lowest & C6 highest
#Team - This company consists of 5 different teams/department
#Raise - # of raises received
#Term..Type - Among employees who left, voluntary leaves are shown as "Quit" & involuntary leaves as "Terminated". Active employees are left blank.
#Dist..To.work - Distance to work from home in km
#I believe the rest are pretty self explantory
head(hr)
## S.EMP. Title C.LEVEL Team Job.Level
## 1 60060523 Senior Art Director C3 Creative Services Supervisor
## 2 60071662 Asso. Creative Director C3 Creative Services Supervisor
## 3 60072318 Sr. Account Executive C2 Account Services Supervisor
## 4 60072838 Engagement Supervisor C3 Account Services Supervisor
## 5 60081603 Junior Art Director C2 Production Associate
## 6 60114364 Account Coordinator C1 Account Services Associate
## BEGIN.SALARY Raise Hire H.Year Termination T.Year TermType
## 1 70000 1 20060821 2006 20110831 2011 Quit
## 2 80000 2 20070522 2007 20110520 2011 Quit
## 3 70000 1 20070816 2007 20120323 2012 Quit
## 4 75000 - 20070910 2007 20120120 2010 Quit
## 5 35000 2 20080520 2008 20140919 2014 Quit
## 6 35000 2 20110501 2011 20140314 2014 Quit
## Status DOB SEX Education
## 1 Terminated 19740928 M Bachelors Degree
## 2 Terminated 19660123 F Bachelors Degree
## 3 Terminated 19800925 M Bachelors Degree
## 4 Terminated 19731212 F Bachelors Degree
## 5 Terminated 19811224 M Bachelors Degree
## 6 Terminated 19830126 F College/Diploma/Associate
## Major DistToWork
## 1 Advertising & Graphic Design 15.0
## 2 English & Religion 40.0
## 3 25.0
## 4 Sociology 20.0
## 5 Graphic Design 13.3
## 6 24.5
summary(hr)
## S.EMP. Title C.LEVEL
## Min. :20001890 Account Executive : 9 C1 :20
## 1st Qu.:60124826 ART DIRECTOR : 8 C2 :42
## Median :60140323 ACCOUNT SUPERVISOR : 5 C3 :40
## Mean :60611386 WRITER : 5 C4 : 9
## 3rd Qu.:60160768 Account Coordinator: 4 C5 :14
## Max. :90150168 Account Supervisor : 4 C6 : 2
## (Other) :92
## Team Job.Level BEGIN.SALARY
## Account Services :47 Associate :67 Min. : 30000
## Corporate Mgmt : 8 Manager : 7 1st Qu.: 45000
## Creative Services:38 Representative Officer: 2 Median : 60000
## Planning :19 Sr.Manager :22 Mean : 72367
## Production :15 Supervisor :29 3rd Qu.: 82500
## Max. :250000
##
## Raise Hire H.Year Termination
## - :76 Min. :20060515 Min. :2006 Min. :20110520
## 1 :31 1st Qu.:20120410 1st Qu.:2012 1st Qu.:20140228
## 2 :15 Median :20140310 Median :2014 Median :20150329
## 3 : 4 Mean :20137445 Mean :2014 Mean :20151022
## 4 : 1 3rd Qu.:20160110 3rd Qu.:2016 3rd Qu.:20170527
## Max. :20190612 Max. :2019 Max. :20190630
## NA's :17
## T.Year TermType Status DOB SEX
## Min. :2010 :17 Active : 17 Min. :19601005 F:66
## 1st Qu.:2014 Quit :78 Terminated:110 1st Qu.:19750516 M:61
## Median :2015 Terminated:32 Median :19830205
## Mean :2015 Mean :19816494
## 3rd Qu.:2017 3rd Qu.:19871206
## Max. :2019 Max. :19950216
## NA's :17
## Education Major
## Bachelors Degree :85 Advertising :21
## College/Diploma/Associate:33 :10
## Masters Degree : 8 Design and Applied Arts: 9
## MBA : 1 Marketing&Sales : 8
## Business Communications: 4
## Commerce : 4
## (Other) :71
## DistToWork
## Min. : 0.40
## 1st Qu.: 5.00
## Median : 8.80
## Mean : 16.74
## 3rd Qu.: 20.30
## Max. :111.00
##
str(hr)
## 'data.frame': 127 obs. of 18 variables:
## $ S.EMP. : int 60060523 60071662 60072318 60072838 60081603 60114364 60115474 60116117 60117843 60117854 ...
## $ Title : Factor w/ 77 levels "Account Coordinator",..: 59 18 64 38 50 1 52 70 49 24 ...
## $ C.LEVEL : Factor w/ 6 levels "C1","C2","C3",..: 3 3 2 3 2 1 3 3 2 2 ...
## $ Team : Factor w/ 5 levels "Account Services",..: 3 3 1 1 5 1 4 5 3 3 ...
## $ Job.Level : Factor w/ 5 levels "Associate","Manager",..: 5 5 5 5 1 1 5 5 1 1 ...
## $ BEGIN.SALARY: int 70000 80000 70000 75000 35000 35000 50000 75000 45000 45000 ...
## $ Raise : Factor w/ 5 levels " - ","1","2",..: 2 3 2 1 3 3 1 2 1 1 ...
## $ Hire : int 20060821 20070522 20070816 20070910 20080520 20110501 20110725 20110822 20111216 20111216 ...
## $ H.Year : int 2006 2007 2007 2007 2008 2011 2011 2011 2011 2011 ...
## $ Termination : int 20110831 20110520 20120323 20120120 20140919 20140314 20120328 20140815 20120413 20130405 ...
## $ T.Year : int 2011 2011 2012 2010 2014 2014 2012 2014 2012 2013 ...
## $ TermType : Factor w/ 3 levels "","Quit","Terminated": 2 2 2 2 2 2 2 2 2 2 ...
## $ Status : Factor w/ 2 levels "Active","Terminated": 2 2 2 2 2 2 2 2 2 2 ...
## $ DOB : int 19740928 19660123 19800925 19731212 19811224 19830126 19830712 19721010 19790830 19830218 ...
## $ SEX : Factor w/ 2 levels "F","M": 2 1 2 1 2 1 1 2 1 2 ...
## $ Education : Factor w/ 4 levels "Bachelors Degree",..: 1 1 1 1 1 2 1 1 1 1 ...
## $ Major : Factor w/ 51 levels "","Accounting and Finance",..: 4 29 1 50 34 1 43 33 20 30 ...
## $ DistToWork : num 15 40 25 20 13.3 24.5 23 45.3 8 5.2 ...
#Plotting a bar chart to see if there's a particular group of job level that stands out
#Please note that I am using Plotly library to take advantage of its interactive functionality (i.e. hover-over data display, etc)
#While each level can be considered as 2-3 years experience accumulatively, C2 & C3 level seem to be harder to be retained (most "Quit"" job levels)
#Among employees terminated, again C2 & C3 level take up more than 50%. I also think it's interesting that C5 level takes sizable portion in "terminated" group.
a<-ggplot(hr, aes(x=TermType, fill=C.LEVEL))+
geom_bar()
ggplotly(a)
#C5 level has avg. 158K salary which is a significant investment from Employer's perspective. Then the fact they are one of the biggest groups who get "terminated" (not quit) by the company may suggests that there is an issue with hiring process.
aggregate(hr$BEGIN.SALARY, by=list(hr$C.LEVEL), FUN=mean)
## Group.1 x
## 1 C1 36275.00
## 2 C2 52430.95
## 3 C3 72075.00
## 4 C4 91666.67
## 5 C5 158571.43
## 6 C6 167500.00
#Checking out the same employee status by department.
#Account and Creative teams are advertising's bread and butter type of roles.
#We seem to have more difficulty retaining Account team. Quit > Terminated
#While retention is still problematic with Creative (Quit 23), it is the department where we are having most problem hiring the right resource (Terminated 11)
b <-ggplot(hr, aes(x=TermType, fill=Team))+
geom_bar()
ggplotly(b)
#Checking # of hires by year and department. There must be a reason why # of hires jumped so greatly after 2011. This also shows that we are in constant need for Account and Creative Services then Planning and Production needs have grown after 2010.
ggplot(hr, aes(x=H.Year, fill=Team))+
geom_bar()
#Changing HireDate, TerminationDate and BirthDate to Date format for further data manipulation.
hdate <- as.character(hr$Hire)
tdate <- as.character(hr$Termination)
bdate <- as.character(hr$DOB)
hr$hdate = as.Date(hdate, format="%Y%m%d")
hr$tdate = as.Date(tdate, format="%Y%m%d")
hr$bdate = as.Date(bdate, format="%Y%m%d")
head(hr)
## S.EMP. Title C.LEVEL Team Job.Level
## 1 60060523 Senior Art Director C3 Creative Services Supervisor
## 2 60071662 Asso. Creative Director C3 Creative Services Supervisor
## 3 60072318 Sr. Account Executive C2 Account Services Supervisor
## 4 60072838 Engagement Supervisor C3 Account Services Supervisor
## 5 60081603 Junior Art Director C2 Production Associate
## 6 60114364 Account Coordinator C1 Account Services Associate
## BEGIN.SALARY Raise Hire H.Year Termination T.Year TermType
## 1 70000 1 20060821 2006 20110831 2011 Quit
## 2 80000 2 20070522 2007 20110520 2011 Quit
## 3 70000 1 20070816 2007 20120323 2012 Quit
## 4 75000 - 20070910 2007 20120120 2010 Quit
## 5 35000 2 20080520 2008 20140919 2014 Quit
## 6 35000 2 20110501 2011 20140314 2014 Quit
## Status DOB SEX Education
## 1 Terminated 19740928 M Bachelors Degree
## 2 Terminated 19660123 F Bachelors Degree
## 3 Terminated 19800925 M Bachelors Degree
## 4 Terminated 19731212 F Bachelors Degree
## 5 Terminated 19811224 M Bachelors Degree
## 6 Terminated 19830126 F College/Diploma/Associate
## Major DistToWork hdate tdate bdate
## 1 Advertising & Graphic Design 15.0 2006-08-21 2011-08-31 1974-09-28
## 2 English & Religion 40.0 2007-05-22 2011-05-20 1966-01-23
## 3 25.0 2007-08-16 2012-03-23 1980-09-25
## 4 Sociology 20.0 2007-09-10 2012-01-20 1973-12-12
## 5 Graphic Design 13.3 2008-05-20 2014-09-19 1981-12-24
## 6 24.5 2011-05-01 2014-03-14 1983-01-26
#Figuring out each employee's Age at hire and Age at termination.
hr$hireage <- as.integer(round((hr$hdate-hr$bdate)/365, digit=0))
hr$termage <- as.integer(round((hr$tdate-hr$bdate)/365, digit=0))
head(hr)
## S.EMP. Title C.LEVEL Team Job.Level
## 1 60060523 Senior Art Director C3 Creative Services Supervisor
## 2 60071662 Asso. Creative Director C3 Creative Services Supervisor
## 3 60072318 Sr. Account Executive C2 Account Services Supervisor
## 4 60072838 Engagement Supervisor C3 Account Services Supervisor
## 5 60081603 Junior Art Director C2 Production Associate
## 6 60114364 Account Coordinator C1 Account Services Associate
## BEGIN.SALARY Raise Hire H.Year Termination T.Year TermType
## 1 70000 1 20060821 2006 20110831 2011 Quit
## 2 80000 2 20070522 2007 20110520 2011 Quit
## 3 70000 1 20070816 2007 20120323 2012 Quit
## 4 75000 - 20070910 2007 20120120 2010 Quit
## 5 35000 2 20080520 2008 20140919 2014 Quit
## 6 35000 2 20110501 2011 20140314 2014 Quit
## Status DOB SEX Education
## 1 Terminated 19740928 M Bachelors Degree
## 2 Terminated 19660123 F Bachelors Degree
## 3 Terminated 19800925 M Bachelors Degree
## 4 Terminated 19731212 F Bachelors Degree
## 5 Terminated 19811224 M Bachelors Degree
## 6 Terminated 19830126 F College/Diploma/Associate
## Major DistToWork hdate tdate bdate
## 1 Advertising & Graphic Design 15.0 2006-08-21 2011-08-31 1974-09-28
## 2 English & Religion 40.0 2007-05-22 2011-05-20 1966-01-23
## 3 25.0 2007-08-16 2012-03-23 1980-09-25
## 4 Sociology 20.0 2007-09-10 2012-01-20 1973-12-12
## 5 Graphic Design 13.3 2008-05-20 2014-09-19 1981-12-24
## 6 24.5 2011-05-01 2014-03-14 1983-01-26
## hireage termage
## 1 32 37
## 2 41 45
## 3 27 32
## 4 34 38
## 5 26 33
## 6 28 31
#Plotting a graph to see if this company is more appealing to certain age group as a career opportunity. We seem to have more new hires in younger or equal to 30 group.
d<-ggplot(hr, aes(x=hireage))+
geom_bar()
ggplotly(d)
#Checking in what age people leave the company the most. Please note that active 17 employees are not included in this plot. It is most motable around 30. Please note that this does not differentiate "Quit" and "Terminated" status. Further analysis is required.
ggplot(hr, aes(x=termage))+
geom_bar()
## Warning: Removed 17 rows containing non-finite values (stat_count).
# Two charts that differentiate Quit and Terminiated.
hr %>% filter(TermType=="Quit" | TermType=="Terminated")%>%
ggplot(., aes(x=termage))+
geom_bar()+
facet_wrap(~TermType)
#Figuring out number of years employees stayed in the company.
hr$tenure <- as.integer(round((hr$tdate-hr$hdate)/365, digit=0))
#Maybe employees would feel that they have had enough and start looking after certain number of years. Also employer could scrutinize employee's performance for a certain period, then maybe the level of scrutiny gets reduced. Below plot suggests that change in staff mostly happens within 1st 3 years.
e<-ggplot(hr, aes(x=TermType, fill=as.factor(tenure)))+
geom_bar()
ggplotly(e)
#Plot by gender. I don't believe this provides much insight. Might be an good illustration of gender equality at workplace.
g<-ggplot(hr, aes(x=TermType, fill=SEX))+
geom_bar()
ggplotly(g)
#Although this analysis is mostly based on employee's personal(objective), # of raises received throughout the tenure is one subjective factor that can illustrate employee's performance.
#While it is obvious that employees without getting recognition (0 raise) are prone to retention risk (both voluntarily and involuntarily), the analysis could further develop with employees with 1 or more raises.
h<-ggplot(hr, aes(x=TermType, fill=Raise))+
geom_bar()
ggplotly(h)
#Plot created to find correlation between employees' Distance to work and Tenure.
ggplotly(ggplot(hr, aes(x=tenure, y=DistToWork))+
geom_jitter())
#Now some relationship between some variables and employee retention has been explored, I would also like to explore some other visualization tools for stakeholders.
# Below is explored in the beginning but testing it in a different format to determine what is easier to understand.
ggplot(hr, aes(x=Team, fill=Team))+
geom_bar()+
facet_wrap(~TermType)+
theme(axis.text.x = element_text(angle = 90, hjust = 1))
#Same exercise as above. Exploring different format.
p<-ggplot(hr, aes(x=Raise, fill=Raise))+
geom_bar()+
facet_wrap(~TermType)
ggplotly(p)
#Exploring HighCharter library for better visualization of analysis, in additioned to enhanced information delivery.
plot_ly(hr, y=hr$tenure, color=hr$Team, type="box")
## Warning: Ignoring 17 observations
#Exploring HighCharter library for better visualization of analysis part 2
hchart(hr, "scatter", hcaes(x=as.integer(BEGIN.SALARY), y=tenure, group=Team))
## Warning: `parse_quosure()` is deprecated as of rlang 0.2.0.
## Please use `parse_quo()` instead.
## This warning is displayed once per session.
#A different sample utilizing collapsibleTree to add interactive function. Might be beneficial when making a presentation.
library(collapsibleTree)
hr %>%
group_by(Team, Education) %>%
summarize('TeamNum' = n())%>%
collapsibleTreeSummary(hr,
hierarchy = c("Team", "Education"),
root = "HR",
width = 800,
attribute = "TeamNum",
collapsed = F,
linkLength = 200
)